CN113288092B - Blood pressure detection method and terminal for extracting pulse wave based on video - Google Patents
Blood pressure detection method and terminal for extracting pulse wave based on video Download PDFInfo
- Publication number
- CN113288092B CN113288092B CN202110491843.6A CN202110491843A CN113288092B CN 113288092 B CN113288092 B CN 113288092B CN 202110491843 A CN202110491843 A CN 202110491843A CN 113288092 B CN113288092 B CN 113288092B
- Authority
- CN
- China
- Prior art keywords
- blood pressure
- pulse wave
- video
- neck
- pulsation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0059—Measuring for diagnostic purposes; Identification of persons using light, e.g. diagnosis by transillumination, diascopy, fluorescence
- A61B5/0062—Arrangements for scanning
- A61B5/0064—Body surface scanning
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0002—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network
- A61B5/0015—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by features of the telemetry system
- A61B5/0024—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by features of the telemetry system for multiple sensor units attached to the patient, e.g. using a body or personal area network
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/021—Measuring pressure in heart or blood vessels
- A61B5/02108—Measuring pressure in heart or blood vessels from analysis of pulse wave characteristics
- A61B5/02125—Measuring pressure in heart or blood vessels from analysis of pulse wave characteristics of pulse wave propagation time
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Veterinary Medicine (AREA)
- Surgery (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Cardiology (AREA)
- Computer Networks & Wireless Communication (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Vascular Medicine (AREA)
- Physiology (AREA)
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The invention provides a blood pressure detection method and a terminal based on video extraction pulse waves, and the method comprises the following steps: initializing parameters of a nonlinear blood pressure calculation model; shooting the video of the pulsating region of the neck and the facial artery of a subject by adopting a camera; amplifying weak neck artery pulse signals and facial artery pulse signals in the two areas by using a motion amplification technology; respectively selecting neck and facial artery pulsation point areas with obvious pulsation and calculating to obtain pulse wave conduction time between the two point areas; and calculating to obtain the average blood pressure value of the subject in the shooting time by adopting a nonlinear blood pressure calculation model. According to the invention, the neck and facial artery areas are collected by the camera, the pulse wave conduction time of the neck and facial artery pulsation points is obtained through signal amplification and extraction and calculation, and finally the blood pressure value is obtained through calculation by combining with a nonlinear blood pressure calculation model, so that noninvasive, non-contact and non-sensory blood pressure measurement is realized while time delay errors are avoided.
Description
Technical Field
The invention relates to the field of biomedicine and digital image processing, in particular to a blood pressure detection method and a terminal based on video extraction pulse wave.
Background
Blood pressure is a very important human physiological parameter, and most of blood pressure detection technologies and instruments in the market are based on invasive, contact and discontinuous applications. However, the current clinical-based invasive blood pressure detection has potential harm to human body, is not easy to operate, and has complicated steps; the sphygmomanometer used for daily health assessment and hospital physical examination belongs to a contact measuring instrument, the discomfort to a measurer can be increased by a measuring mode of cuff pressurization, the convenience is low, and the change of blood pressure cannot be continuously monitored.
Because the conduction time of the pulse wave has a certain correlation with the blood pressure, most of the existing methods for non-invasively detecting the blood pressure value are based on obtaining the pulse wave of a human body and measuring the conduction time of the pulse wave, when the pulse wave is transmitted from the heart to an arterial system, the pulse wave is influenced not only by the heart, but also by various physiological factors such as vascular resistance, vascular wall elasticity, vascular radius change, blood viscosity and the like flowing through various levels of arteries and branches, so that the pulse wave contains abundant physiological and pathological information of the cardiovascular system. The current methods for obtaining pulse waves can be divided into two categories: photoplethysmography (PPG) and imaging photoplethysmography (IPPG). The PPG technology irradiates skin tissues of a human body through light beams, the light beams are transmitted to a photoelectric receiver in a reflection or projection mode, and optical signals are converted into electric signals according to the change of the light intensity received by the photoelectric receiver to obtain pulse wave signals; this technique requires the human body to be in close contact with the sensor and to be fixed in a certain measured area, which is unavoidable for human discomfort. IPPG technology uses a camera to capture the change of skin color and then extract pulse waves, and is widely used due to its non-contact characteristic. However, when the PPG or IPPG is used to obtain the pulse wave transit time, it is necessary to detect the blood pressure value by combining the acquisition of ECG (electrocardiogram) signals, and the pulse wave transit time calculated by using the R wave peak of the ECG as the starting point actually includes two parts: isovolumetric contraction period and true pulse wave transit time, so that the pulse wave transit time measured by using the traditional technology has delay errors.
At present, two pulse wave signal acquisition methods are usually adopted for measuring the pulse wave conduction time, for example, a PPG sensor is placed on a brachial artery and a finger artery to acquire a pulse wave signal and pulse wave conduction time, an ECG signal is not required to be acquired in the method, delay errors caused by the isovolumetric contraction period of the heart are avoided, the limitation that the sensor is in contact with a human body cannot be avoided, and the distance between the sensors is changed due to the movement of the body, so that the influence effect is large.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method and the terminal for detecting the blood pressure based on the video extraction pulse wave are provided, and the defects of the extraction of the pulse wave in the prior art are overcome while the noninvasive, non-contact and non-inductive blood pressure detection is realized.
In order to solve the technical problems, the invention adopts the technical scheme that:
a blood pressure detection method based on video extraction pulse waves comprises the following steps:
s1, initializing parameters of the nonlinear blood pressure calculation model;
s2, initializing a camera, wherein the camera shoots a face video of the subject, and intercepts a neck artery pulsation region video and a face artery pulsation region video;
s3, amplifying the neck artery pulsation region video and the facial artery pulsation region video by using a motion amplification technology respectively, and amplifying weak neck artery pulsation signals and facial artery pulsation signals in the videos;
s4, respectively selecting point areas with obvious pulsation in the amplified neck artery pulsation area video and the amplified facial artery pulsation area video as a neck artery pulsation point area and a facial artery pulsation point area, and calculating to obtain the pulse wave conduction time between the two pulsation point areas;
and S5, calculating to obtain the average blood pressure value of the subject in the shooting time by combining the pulse wave conduction time, the parameters and the nonlinear blood pressure calculation model.
In order to solve the technical problem, the invention adopts another technical scheme as follows:
a blood pressure detection terminal for extracting pulse waves based on videos comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the blood pressure detection method for extracting pulse waves based on videos when executing the computer program.
The invention has the beneficial effects that: the invention provides a blood pressure detection method and a terminal for extracting pulse waves based on videos.
Drawings
FIG. 1 is a schematic diagram of a method for detecting blood pressure based on video extraction of pulse waves;
FIG. 2 is a main flow chart of a blood pressure detection method based on video extraction of pulse waves;
FIG. 3 is a flowchart of a blood pressure detecting method according to the second embodiment;
FIG. 4 is a flow chart of a simultaneous measurement method for fitting parameters according to a second embodiment;
fig. 5 is a flowchart of a camera according to a first embodiment for capturing and acquiring a regional video;
FIG. 6 is a flowchart of signal amplification and pulse wave extraction according to the first embodiment;
FIG. 7 is a schematic diagram showing the selection of the carotid artery pulsation point and the facial artery pulsation point;
FIG. 8 is a waveform of the neck pulse wave and the face pulse wave;
fig. 9 is a block diagram of a blood pressure detecting terminal for extracting pulse waves based on video.
Description of the reference symbols:
1. a computer; 2. a camera; 3. no stroboscopic light source; 4. an adjustable support; 5. a subject; 10. a blood pressure detection terminal for extracting pulse waves based on video; 20. a memory; 30. a processor.
Detailed Description
In order to explain technical contents, achieved objects, and effects of the present invention in detail, the following description is made with reference to the accompanying drawings in combination with the embodiments.
Referring to fig. 1 to 8, a blood pressure detecting method for extracting pulse wave based on video includes the steps of:
s1, initializing parameters of the nonlinear blood pressure calculation model;
s2, initializing a camera, wherein the camera shoots a face video of the subject, and intercepts a neck artery pulsation region video and a face artery pulsation region video;
s3, amplifying the neck artery pulsation region video and the facial artery pulsation region video by using a motion amplification technology respectively, and amplifying weak neck artery pulsation signals and facial artery pulsation signals in the videos;
s4, respectively selecting point areas with obvious pulsation in the amplified neck artery pulsation area video and the amplified facial artery pulsation area video as a neck artery pulsation point area and a facial artery pulsation point area, and calculating to obtain the pulse wave conduction time between the two pulsation point areas;
and S5, calculating to obtain the average blood pressure value of the subject in the shooting time by combining the pulse wave conduction time, the parameters and the nonlinear blood pressure calculation model.
As can be seen from the above description, the beneficial effects of the present invention are: the method comprises the steps of collecting a neck region and a face region of a subject through a camera, amplifying weak neck artery pulsating signals and face artery pulsating signals in the regions by using a motion amplification technology, extracting the neck artery pulsating point regions and the face artery pulsating point regions with obvious pulsation, calculating to obtain pulse wave conduction time between the two pulsating point regions, combining pre-fitted parameters and a nonlinear blood pressure calculation model to obtain an average blood pressure value of the subject within the shooting time, and realizing noninvasive, non-contact and non-sensory blood pressure measurement while avoiding time delay errors.
Further, the step S1 is specifically:
and carrying out multiple experimental fitting on a plurality of subjects in advance to obtain the parameters of the nonlinear blood pressure calculation model.
Further, the step S1 is specifically:
obtaining the parameter by using a synchronous measurement method, wherein the synchronous measurement method specifically comprises the following steps:
and executing the steps S2 to S4 to obtain the pulse wave conduction time, simultaneously measuring by using an electronic sphygmomanometer to obtain a real-time blood pressure value, repeating the steps to obtain a plurality of groups of pulse wave conduction time and the real-time blood pressure value, and combining the nonlinear blood pressure calculation model to obtain the parameters.
According to the description, the parameters of the nonlinear blood pressure calculation model are obtained through multiple times of experimental fitting by multiple experimenters, but the health conditions of arterial blood vessels of different people are different, so that a subject can select to directly adopt the parameters or select to self-fit personalized parameters, only a synchronous measurement method is needed, namely, the electronic sphygmomanometer is used for obtaining real-time blood pressure values while a video camera shoots neck and face videos of the subject and obtains pulse wave conduction time through motion amplification technology and calculation, and therefore the personalized parameters of the subject are obtained through fitting according to the nonlinear blood pressure calculation model, and personalized parameter initialization is achieved.
Further, the initializing the camera in step S2 is specifically: the camera adopts a high frame rate mode, and the shooting time of the camera is [5s,10s ].
As can be seen from the above description, for people of different ages, different body parts, whether vascular diseases exist or not, different gender, different pulse wave conduction velocity and the like, the conduction velocity of the pulse wave is approximately between 9-19m/s, the measured artery is selected as the neck artery and the facial artery of the subject, the pulse point distance between the two points is close, the pulse wave conduction velocity is in the millisecond level, and the high frame rate mode of the camera can ensure the measurement accuracy; meanwhile, stable regional video can be obtained within a certain shooting time, and the measurement accuracy is further improved.
Further, the step S2 further includes:
performing feature point detection and image tilt correction on the photographed video;
if the ambient light is insufficient or the light changes continuously, a non-stroboscopic light source is adopted to continuously irradiate the shot part of the subject.
According to the description, the face shaking interference can be eliminated through the feature point detection and the image inclination correction, and the distortion influence on the amplification of the video picture motion signal caused by the face shaking is avoided; the non-stroboscopic light source is adopted when the ambient light is insufficient or the light continuously changes, so that the influence of the change of the light intensity on the pulse wave oscillogram can be avoided, and the measurement accuracy is further improved.
Further, the step S3 further includes:
the motion amplification technique is a fast phase amplification technique based on the Riesz pyramid.
As can be seen from the above description, the pulse signal of the artery is extremely small, and the small pulse signal can be effectively amplified by using the fast phase amplification technology based on the Riesz pyramid.
In step S4, the pulse transit time between the two pulse point areas calculated and obtained is specifically:
respectively taking the middle points of the neck artery pulsating point area and the facial artery pulsating point area as centers, taking the area of 3 multiplied by 3 and solving a brightness average value, obtaining a neck pulse wave and a facial pulse wave according to the change of the brightness average value along with time, and calculating the peak value difference of the adjacent wave crests of the neck pulse wave and the facial pulse wave in the same cardiac cycle;
the peak differences are calculated over a number of different cardiac cycles and a weighted average is found as the pulse transit time.
It can be known from the above description that the pulsation after signal amplification is clearly seen by naked eyes, but because the pulsation intensity of the artery of different people is different, a pulsation point area with obvious pulsation in the area needs to be selected, in order to improve the accuracy of the obtained pulse wave, a 3 × 3 area is selected as the pulsation point area, the change of the brightness average value is calculated, and meanwhile, the average value of a plurality of groups of peak value differences is calculated as the pulse wave propagation time, so that the measurement accuracy can be further improved.
Further, the obtaining a neck pulse wave and a face pulse wave according to the variation of the brightness average value with time, and calculating a peak difference between adjacent peaks of the neck pulse wave and the face pulse wave in the same cardiac cycle further includes:
and filtering the neck pulse wave and the face pulse wave by adopting a blind source separation technology.
According to the above description, the obtained pulse wave is mixed with the pulse wave signal of the region of interest, the self-movement signal of the human body and the noise signal, and the more accurate pulse wave can be obtained after the filtering processing is performed by the blind source separation technology, so that the measurement accuracy is further improved.
In an embodiment, the formula of the nonlinear blood pressure calculation model is:
p is the mean blood pressure value, T PTT A, b, c, and d are the parameters.
From the above description, it can be known that the parameter value can be obtained by fitting the real-time blood pressure value obtained by measuring with the electronic sphygmomanometer in combination with the formula of the nonlinear blood pressure calculation model in the embodiments, and meanwhile, the blood pressure value can also be obtained by fitting the formula of the nonlinear blood pressure calculation model and the parameter obtained in advance according to the pulse wave propagation time obtained by measuring.
Referring to fig. 9, a blood pressure detecting terminal for extracting pulse wave based on video includes a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor implements a blood pressure detecting method for extracting pulse wave based on video according to any one of claims 1 to 9 when executing the computer program.
As can be seen from the above description, the beneficial effects of the present invention are: based on the same technical concept, the blood pressure detection method based on video extraction pulse waves is matched, a blood pressure detection terminal based on video extraction pulse waves is provided, a neck region and a face region of a subject are collected through a camera, weak neck artery pulse signals and face artery pulse signals in the regions are amplified by using a motion amplification technology, a neck artery pulse point region and a face artery pulse point region with obvious pulse are extracted, pulse wave conduction time between the two pulse point regions is obtained through calculation, an average blood pressure value of the subject in the shooting time is obtained through combination of pre-fitted parameters and a nonlinear blood pressure calculation model, and noninvasive, non-contact and non-sensory blood pressure measurement is achieved while time delay errors are avoided.
Referring to fig. 1, fig. 2, fig. 5, fig. 6, fig. 7 and fig. 8, a first embodiment of the present invention is:
a blood pressure detection method based on video extraction pulse waves is disclosed, as shown in figure 1, a system adopted by the detection method is composed of a computer 1, a camera 2 and a non-stroboscopic light source 3, the camera 2 is adjusted to be positioned in front of the neck and the head of a subject 5 through an adjustable support 4, and facial videos of the subject are collected through the camera 2. As shown in fig. 2, the blood pressure monitoring method includes the following steps:
and S1, initializing parameters of the nonlinear blood pressure calculation model.
S2, initializing a camera, shooting a facial video of the subject by the camera, and intercepting a neck artery pulsation region video and a facial artery pulsation region video;
the pulse wave propagation velocity is approximately 9-19m/s because the pulse wave propagation velocity ranges are different for people of different ages, different body parts, whether vascular diseases exist, sex difference, pulse wave propagation velocity and the like. In this embodiment, as shown in fig. 7, the measured artery is selected as the neck artery and facial artery of the subject, the distance between the two points is relatively short, and the pulse wave propagation speed is in the order of milliseconds. In this embodiment, the video camera selects a high frame rate mode with a frame rate of 240Hz and a resolution of 1280 × 720, and in other equivalent embodiments, different frame rate modes may be selected according to actual needs.
As shown in fig. 5, in order to avoid the distortion effect of the face shake on the amplification of the video picture motion signal, the feature point detection and the image tilt correction are performed on the shot video to eliminate the face shake interference; when the ambient light is insufficient or the light continuously changes, a non-stroboscopic light source is required to continuously irradiate the shot part of the subject, so that the influence of the change of the light intensity on the pulse wave oscillogram is avoided; meanwhile, the shooting time is required to be kept to be 5-10s during shooting, stable neck and face region videos are guaranteed to be obtained, and due to the fact that different subjects have different beating regions, the neck and face region videos need to be selected and cut manually, and the measuring accuracy is further improved.
S3, amplifying the neck artery pulsation region video and the facial artery pulsation region video by using a motion amplification technology respectively, and amplifying weak neck artery pulsation signals and facial artery pulsation signals in the videos;
as shown in fig. 6, in the present embodiment, the computer performs signal amplification processing on the clipped neck region video and face region video, and amplifies a minute pulse signal by using a fast phase amplification technique based on the Riesz pyramid.
S4, respectively selecting point areas with obvious pulsation in the amplified neck artery pulsation area video and facial artery pulsation area video as a neck artery pulsation point area and a facial artery pulsation point area, and calculating to obtain the pulse wave conduction time between the two pulsation point areas;
the pulsation after signal amplification is clearly visible to naked eyes, but because the pulsation intensity of the artery of different people is different, a pulsation point area with obvious pulsation in the area needs to be selected. In order to further improve the accuracy of the acquired pulse wave, a 3 × 3 area is selected as a pulse point area, the change of the brightness average value is calculated, and the average value of a plurality of groups of peak value differences is calculated as the pulse wave propagation time.
Wherein the pulse wave conduction time between the two pulsation points obtained by calculation is specifically as follows:
respectively taking the middle point of a neck artery pulsation point area and a facial artery pulsation point area as the center, taking a 3 multiplied by 3 area and solving the brightness average value, obtaining a neck pulse wave and a facial pulse wave according to the change of the brightness average value along with time, calculating the peak value difference of the adjacent wave crests of the neck pulse wave and the facial pulse wave in the same cardiac cycle, calculating the peak value difference in a plurality of different cardiac cycles, and obtaining the weighted average value as the pulse wave conduction time.
In this embodiment, as shown in fig. 7, after the cervical artery pulsation point B and the facial artery pulsation point a are selected, the spatiotemporal slices are obtained respectively with the two points as the centers, that is, the areas of 3 × 3 are respectively taken and the brightness average value is obtained, and then the change of the brightness average value with time is recorded as the original pulse wave signal x of the two areas 1 (t)。
Selecting any edge of the previously selected 3 × 3 regions as a neighboring edge, selecting the bottom edge and the right edge in this embodiment, selecting two 3 × 3 regions, calculating the average brightness values in the regions, and recording the change of the average brightness values of the two regions with time as the original pulse wave signal x 2 (t) and x 3 (t)。
Because the original pulse wave signal is mixed with the pulse wave signal of the interested area, the self-movement signal of the human body and the noise signal, the original pulse wave needs to be filtered by adopting a blind source separation technology.
According to the blind source separation idea, the acquired original pulse wave signal x is subjected to 1 (t)、x 2 (t) and x 3 (t), also called observation signal, applies Independent Component Analysis (ICA) to obtain the pulse wave signal S of the region of interest 1 (t) signal of the body' S own motion S 2 (t) and a noise signal S 3 (t), the source signal.
The ICA model algorithm assumes linear mixing of source signals when observing signals to obtain a relation:
the mixing equation can be used to express:
x(t)＝As(t)
wherein x (t) ═ x 1 (t),x 2 (t),x 3 (t)] T ，s(t)＝[s 1 (t),s 2 (t),s 3 (t)] T A is a 3 × 3 mixing matrix, a ij Are the mixing matrix coefficients. The goal of the ICA algorithm is to find a mixing matrix W, the inverse of which -T Approximating the original mixing matrix A, the output is an estimate of the vector s (t) containing the potential source signals, i.e.
As shown in the following formula:
after filtering the pulse wave signal, the human body self-movement signal and the noise signal of the region of interest, obtaining an accurate pulse wave, obtaining a waveform diagram of a two-path pulse wave extraction result, as shown in fig. 8, calculating peak differences of the two-path pulse waves in a plurality of cardiac cycles, and obtaining an average value of a plurality of groups of peak differences as a pulse wave conduction time T PPT I.e. by
S5, calculating to obtain an average blood pressure value P of the subject in the shooting time by combining the pulse wave conduction time, the parameters and a nonlinear blood pressure calculation model, wherein the nonlinear blood pressure calculation model has the following formula:
in step S1, the parameters a, b, c, and d are obtained by performing multiple experimental fits on multiple subjects in advance. In the present embodiment, the pulse wave conduction is measured by the 60 embodimentsTime T PTT The blood pressure value P is measured by the cuff-pressure electronic sphygmomanometer and then substituted into the nonlinear blood pressure calculation model, and the parameter values a-0.05294, b-335.85374, c-22.9818 and d-191.1235 are obtained by fitting.
Referring to fig. 2 and fig. 3, a second embodiment of the present invention is:
on the basis of the first embodiment, the method for detecting blood pressure based on video extraction pulse wave further includes step S1, specifically: and carrying out multiple experimental fitting on a plurality of subjects in advance to obtain parameters of the nonlinear blood pressure calculation model.
As shown in fig. 2, in this embodiment, this step is an optional step, and in the first embodiment, it has been described that parameters of the nonlinear blood pressure calculation model are obtained by multiple experimental fits by multiple experimenters, but since the health conditions of arterial blood vessels of different people are also different, a subject may choose to directly use the parameters or choose to fit personalized parameters by itself, that is, the above steps are replaced with:
the parameters are obtained by using a synchronous measurement method, as shown in fig. 3, the synchronous measurement method specifically includes:
while the pulse wave propagation time is obtained in steps S2 to S4 in the first embodiment, a real-time blood pressure value is obtained by measuring with an electronic sphygmomanometer, and the steps are repeated to obtain a plurality of groups of pulse wave propagation times and real-time blood pressure values, and parameters are obtained by combining a nonlinear blood pressure calculation model;
the camera shoots the neck and face videos of the testee, the pulse wave conduction time is obtained through the motion amplification technology and calculation, and meanwhile, the electronic sphygmomanometer is adopted to obtain real-time blood pressure numerical values, so that individual personalized parameters of the testee are obtained through fitting according to the nonlinear blood pressure calculation model, and personalized parameter initialization is achieved. The video is shot and processed in the same way as the subsequent steps in the first embodiment, except that an electronic sphygmomanometer is used to record the real-time blood pressure value.
If the elasticity of the blood vessel is kept unchanged, the change of the blood pressure is in direct proportion to the change of the pulse wave conduction time, a linear blood pressure calculation model can be established:
P＝a+bT PPT
the following two non-linear blood pressure calculation equations are proposed in the related research, and are considered to contain the relationship between the elasticity of the blood vessel wall and the change of the radius of the blood vessel and the blood pressure, and the two non-linear blood pressure calculation equations are changed as follows:
P＝aln(T PPT )+b
blood pressure and T PPT 、ln(T PPT ) And (1/T) PPT ) 2 All have different correlations, and the related research proposes to establish a nonlinear multi-modulus blood pressure calculation model, namely a nonlinear blood pressure calculation model formula in the first embodiment:
the values of the parameters a, b, c and d have been obtained by fitting in the first embodiment, and in the present embodiment, the synchronous measurement of the above steps is adopted, i.e. the measured pulse wave transit time is combined with the nonlinear blood pressure calculation model to obtain the measured blood pressure P f Simultaneously measuring the real-time blood pressure value P by using the cuff-pressurized electronic sphygmomanometer x . According to another 24 embodiments using simultaneous measurement, P is measured by the Bland-Altman difference method x And P f The blood pressure measuring method can effectively replace a cuff pressure type electronic sphygmomanometer to realize noninvasive, non-contact and non-sensible blood pressure measurement, and the difference absolute value MEAN value is 1.8630mmHg which is less than 5mmHg, and the error is 95.83% which is within the consistency limit.
Referring to fig. 9, a third embodiment of the present invention is:
a blood pressure detecting terminal 10 for extracting pulse waves based on video comprises a memory 20, a processor 30 and a computer program stored on the memory 20 and capable of running on the processor 30, wherein in the embodiment, the steps in any one of the first embodiment or the second embodiment are realized when the processor 30 executes the computer program.
Meanwhile, the invention can also be applied to various places such as medical monitoring, monitoring and the like, the human pulse wave can be extracted through the video shooting and motion amplification technology, and the non-invasive, non-contact and non-inductive blood pressure measurement can be realized by adopting the blood pressure measurement method of the embodiment I or the embodiment.
In summary, according to the blood pressure detection method and the terminal for extracting the pulse wave based on the video, the videos of the neck region and the face region of the subject within a certain time are acquired through the camera, the weak neck artery pulsating signals and face artery pulsating signals in the neck region and the face region are amplified by using the motion amplification technology, the neck artery pulsating point region and the face artery pulsating point region with obvious pulsation are extracted, the pulse wave conduction time between the two pulsating point regions is obtained by calculation, the average blood pressure value of the subject within the shooting time is obtained by combining the pre-fitted parameters and the nonlinear blood pressure calculation model, and the noninvasive, non-contact and non-sensory blood pressure measurement is realized while the time delay error is avoided. The pre-fitted parameters can be obtained by selecting the existing fitting parameters or combining a non-linear blood pressure calculation model through a synchronous measurement method according to the needs of a subject, so that the personalized parameter initialization is realized; the method for extracting the pulse wave by using the motion amplification technology, the filtering technology of blind source separation and the average value of the peak value difference of the two paths of pulses are calculated, so that the accuracy of pulse wave extraction is improved, and the influence of time delay errors on measurement is effectively avoided; meanwhile, stable shooting time, characteristic point detection and image inclination correction are executed, and the condition that no stroboscopic light source is adopted to compensate the insufficient ambient light or the light continuously changes is adopted, so that the stability of a shot picture is ensured, the influence of factors such as human and environment on the waveform change of the pulse wave is avoided, and the measurement accuracy is further improved.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all equivalent changes made by using the contents of the present specification and the drawings, or applied directly or indirectly to the related technical fields, are included in the scope of the present invention.
Claims (8)
1. A blood pressure detecting terminal for extracting pulse wave based on video comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the computer program to realize the following steps:
s1, initializing parameters of the nonlinear blood pressure calculation model;
s2, initializing a camera, wherein the camera shoots a face video of the subject, and intercepts a neck artery pulsation region video and a face artery pulsation region video;
s3, amplifying the neck artery pulsation region video and the facial artery pulsation region video by using a motion amplification technology respectively, and amplifying weak neck artery pulsation signals and facial artery pulsation signals in the videos;
s4, respectively selecting point areas with obvious pulsation in the amplified neck artery pulsation area video and the amplified facial artery pulsation area video as a neck artery pulsation point area and a facial artery pulsation point area, and calculating the pulse wave conduction time between the two pulsation point areas, wherein the specific steps are as follows:
respectively taking the middle point of the neck artery pulsating point region and the facial artery pulsating point region as the center, taking a 3 multiplied by 3 region and calculating the brightness average value, and then the change of the brightness average value along with the time is recorded as the original pulse wave signal x of the two regions 1 (t); using any edge of the previously selected 3 × 3 area as a neighboring edge, selecting two 3 × 3 areas, calculating the brightness average value in the areas, and then the change of the brightness average value in the two areas with time is marked as the original pulse wave signal x 2 (t) and x 3 (t); for the original pulse wave signal x 1 (t)、x 2 (t) and x 3 (t) filtering by adopting a blind source separation technology, which specifically comprises the following steps:
according to blind source separationFor the original acquired pulse wave signal x 1 (t)、x 2 (t) and x 3 (t), also called observation signal, using an independent component analysis algorithm to obtain a pulse wave signal S of the region of interest 1 (t) signal of the body' S own motion S 2 (t) and a noise signal S 3 (t), the source signal;
assuming linear mixing of the source signals by the independent component analysis algorithm when the observation signals are obtained, a relation is obtained:
i＝1,2,3；
the mixing equation can be used to express:
x(t)＝As(t)；
wherein x (t) ═ x 1 (t),x 2 (t),x 3 (t)] T ，s(t)＝[s 1 (t),s 2 (t),s 3 (t)] T A is a 3 × 3 mixing matrix, a ij For the coefficients of the mixing matrix, the objective of the independent component analysis algorithm is to find a mixing matrix W, the inverse W -T Approximating the original mixing matrix A, the output of which is an estimate of the vector s (t) containing the potential source signals, i.e.
As shown in the following formula:
filtering the pulse wave signals of the interested region, the human body self-movement signals and the noise signals to obtain accurate pulse waves, namely obtaining neck pulse waves and face pulse waves by the change of the brightness average values of 3 multiplied by 3 regions respectively selected on the neck artery pulsating point region and the face artery pulsating point region along with time, and calculating the peak value difference of the adjacent wave crests of the neck pulse waves and the face pulse waves in the same cardiac cycle;
calculating the peak differences over a plurality of different cardiac cycles and finding a weighted average as the pulse transit time;
and S5, calculating to obtain the average blood pressure value of the subject in the shooting time by combining the pulse wave conduction time, the parameters and the nonlinear blood pressure calculation model.
2. The terminal for detecting blood pressure based on video extraction pulse wave of claim 1, wherein the processor when executing step S1 specifically comprises:
and carrying out multiple experimental fitting on a plurality of subjects in advance to obtain the parameters of the nonlinear blood pressure calculation model.
3. The terminal for detecting blood pressure based on video extraction pulse wave of claim 1, wherein the processor when executing step S1 specifically comprises:
obtaining the parameter by using a synchronous measurement method, wherein the synchronous measurement method specifically comprises the following steps:
and S2 to S4 are executed to obtain the pulse wave conduction time, meanwhile, an electronic sphygmomanometer is used for measuring to obtain a real-time blood pressure value, the steps are repeated to obtain a plurality of groups of pulse wave conduction time and the real-time blood pressure value, and the parameters are obtained by combining the nonlinear blood pressure calculation model.
4. The blood pressure detecting terminal for extracting pulse wave based on video according to claim 1, wherein the processor, when initializing the camera in step S2, specifically:
the camera selects a high frame rate mode, and the shooting time of the camera is [5s,10s ].
5. The terminal for detecting blood pressure based on video extraction pulse wave of claim 1, wherein the processor further comprises, when executing step S2:
performing feature point detection and image tilt correction on the shot video;
if the ambient light is insufficient or the light changes continuously, a non-stroboscopic light source is adopted to continuously irradiate the shot part of the subject.
6. The blood pressure detecting terminal for extracting pulse wave based on video of claim 1, wherein the motion amplification technique is a fast phase amplification technique based on Riesz pyramid.
7. The blood pressure detecting terminal for extracting pulse wave based on video of claim 1, wherein the processor further comprises, before performing step S4, obtaining the neck pulse wave and the face pulse wave from the time variation of the brightness average value, and calculating the peak difference between the adjacent peaks of the neck pulse wave and the face pulse wave in the same cardiac cycle:
and filtering the neck pulse wave and the face pulse wave by adopting a blind source separation technology.
8. The blood pressure detecting terminal for extracting pulse wave based on video of claim 1, wherein the formula of the non-linear blood pressure calculation model is:
p is the mean blood pressure value, T PTT A, b, c and d are the parameters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110491843.6A CN113288092B (en) | 2021-05-06 | 2021-05-06 | Blood pressure detection method and terminal for extracting pulse wave based on video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110491843.6A CN113288092B (en) | 2021-05-06 | 2021-05-06 | Blood pressure detection method and terminal for extracting pulse wave based on video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113288092A CN113288092A (en) | 2021-08-24 |
| CN113288092B true CN113288092B (en) | 2022-08-19 |
Family
ID=77321750
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110491843.6A Active CN113288092B (en) | 2021-05-06 | 2021-05-06 | Blood pressure detection method and terminal for extracting pulse wave based on video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113288092B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113938622B (en) * | 2021-12-15 | 2022-02-15 | 慕思健康睡眠股份有限公司 | Blood pressure detection device based on asynchronously recorded video and storage medium |
| CN114431846B (en) * | 2022-01-12 | 2023-09-22 | 中国科学院力学研究所 | Non-contact measurement method and visual detection system for blood vessel blood pressure in operation |
| CN114569096B (en) * | 2022-02-28 | 2024-05-24 | 西安交通大学 | Non-contact continuous blood pressure measurement method and system based on video stream |
| CN117058131B (en) * | 2023-10-11 | 2024-03-19 | 深圳市鹰瞳智能技术有限公司 | Method for carrying out visual positioning on facial artery based on high-resolution thermal imaging |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105989357A (en) * | 2016-01-18 | 2016-10-05 | 合肥工业大学 | Human face video processing-based heart rate detection method |
| CN110236508A (en) * | 2019-06-12 | 2019-09-17 | 云南东巴文健康管理有限公司 | A kind of non-invasive blood pressure continuous monitoring method |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105147269B (en) * | 2015-06-16 | 2017-10-13 | 江苏斯坦德利医疗科技有限公司 | A kind of noninvasive continuous BP measurement method |
| US11350825B2 (en) * | 2016-08-25 | 2022-06-07 | Vivonics, Inc. | Contactless system and method for measuring and continuously monitoring arterial blood pressure |
| CN106618537B (en) * | 2016-12-21 | 2020-09-01 | 天津普仁万合信息技术有限公司 | Continuous dynamic blood pressure monitoring device and method based on pulse wave conduction |
-
2021
- 2021-05-06 CN CN202110491843.6A patent/CN113288092B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105989357A (en) * | 2016-01-18 | 2016-10-05 | 合肥工业大学 | Human face video processing-based heart rate detection method |
| CN110236508A (en) * | 2019-06-12 | 2019-09-17 | 云南东巴文健康管理有限公司 | A kind of non-invasive blood pressure continuous monitoring method |
Non-Patent Citations (4)
| Title |
|---|
| 基于IPPG非接触式血压测量方法;孙晟珂等;《信息技术》;20201031;第44卷(第10期);第31-38页 * |
| 基于视频放大的脉搏波提取方法研究;李江山;《中国优秀博硕士学位论文全文数据库(硕士) 医药卫生科技辑》;20200115(第01期);第21-38页 * |
| 基于视频非接触技术的身份识别、追踪及异常预警;黄吉羊;《中国优秀博硕士学位论文全文数据库（硕士） 医药卫生科技辑》;20160715(第07期);第10-12，29-43页 * |
| 脉搏波波速法无创血压测量中多模量血压计算模型研究;顾亚雄等;《中国生物医学工程学报》;20161231;第35卷(第6期);第691-698页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113288092A (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113288092B (en) | Blood pressure detection method and terminal for extracting pulse wave based on video | |
| CN110662480B (en) | Device, system and method for measuring and processing physiological signals of a subject | |
| Jeong et al. | Introducing contactless blood pressure assessment using a high speed video camera | |
| CN108471989B (en) | Device, system and method for generating a photoplethysmographic image carrying vital sign information of a subject | |
| EP3188650B1 (en) | Method and apparatus for the continous estimation of human blood pressure using video images | |
| WO2008069062A1 (en) | Blood flow speed image creating device | |
| Sun et al. | Imaging the cardiovascular pulse | |
| US10694959B2 (en) | Image based blood pressure monitoring method | |
| JP6620999B2 (en) | Biological information measuring device, biological information measuring program, and biological information measuring method | |
| JP5535477B2 (en) | Vascular aging detection system | |
| JP2019097757A5 (en) | ||
| Shirbani et al. | Effect of ambient lighting and skin tone on estimation of heart rate and pulse transit time from video plethysmography | |
| WO2021124071A1 (en) | Improved personal health data collection | |
| Zou et al. | Non-contact blood pressure measurement based on iPPG | |
| Hennersperger et al. | Vascular 3D+ T freehand ultrasound using correlation of doppler and pulse-oximetry data | |
| Volynsky et al. | Blood peripheral circulation assessment method based on combined use of the video-capillaroscopy, imaging photoplethysmography, and electrocardiography | |
| Kaviya et al. | Analysis of i-PPG signals acquired using smartphones for the calculation of pulse transit time and oxygen saturation | |
| Xu et al. | Online continuous measurement of arterial pulse pressure and pressure waveform using ultrasound | |
| CA3023735C (en) | High resolution blood perfusion imaging using a camera and a pulse oximeter | |
| Paul et al. | Waveform analysis for camera-based photoplethysmography imaging | |
| Natarajan et al. | Camera-based blood pressure monitoring | |
| Alhammad | Face detection for pulse rate measurement | |
| Cao et al. | Remote physiological monitoring of neck blood vessels | |
| WO2022182956A1 (en) | Remote monitoring of oxygenation status and blood pulsation within skin tissue | |
| Kamshilin et al. | Video-based measurements of blood pulsations delay in human faces |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |